Research

Here is some of my current and planned research. If you are interested in collaboration or working with me send me an email at majmalazad{@}gmail.com.

Decentralized Collaborative Privacy-Aware Network Management

There will be 50 Billion IoT devices by 2050 and global M2M connections will reach close to one billion by 2020, growing at the rate of 25% from the period 2015 to 2020. The collaboration among devices and machines could provide an effective defense against a different type of security attacks and helps in providing aggregated information about the behavior of the machine in the network. The collaboration can be achieved through the exchange of information among collaborating entities, without disclosing any private information of users and machines to any party or trusted third party. The challenge is to design an effective decentralized collaborative Network Management system where machines or users collaborate without worrying about their privacy. In this research project, the motivation is to achieve the collaboration through the exchange of machine learning models rather than exchanging the data itself. 

______________________________________________________________________________________________________________________________________________

Privacy-Aware Outsourcing of Network Functions to Cloud

As per Cisco recent statistics, the global Internet Protocol (IP) traffic will exceed 1000 exabytes per year by the end of 2016 and will reach 2 zettabytes per year by 2019. The service provider, ISP or network operators are receiving gigabits of data per hour and processing a large network data for meaningful information (such intrusion, QoS, QoE) requires extensive system and network resources that analyzed the network traffic flow. For the meaningful data analytic and storage of data for a considerably longer time, the service providers are outsourcing data and network functions (firewall, QoS estimation, Intrusion detection system, etc.) to the cloud service provider for data storage and a meaningful decision over the data. The outsourcing of data and network functions as a challenge of security and privacy threat towards user’s outsourced data and data provider network functions. The challenge is to exchange the data and network functions to the cloud service provider in such a way that cloud service provider would not be able to learn anything about data provider network functions and data.

______________________________________________________________________________________________________________________________________________

Protection Against Caller-ID Spoofing

Caller ID spoofing is when someone changes the Caller ID to any number that he wants to display to the call recipients. Caller ID is actually meant for providing information to callee about the person calling the callee. There are several techniques and smartphone applications are available that can be used to spoof the identity. It can be used to identify the telemarketers and spammers; however, spammers, telemarketers, and prank callers are spoofing identities of legitimate users to circumvent the detection system and fraud user by pretending to be a legitimate entity. It is important to have a system that can identify the users hiding their true identity. The challenge in this regard is to two-fold. 1) Making decision during the call setup phase, and 2) making a decision about the identity of the caller without involving caller and the callee for certain response messages. The call setup messages can provide complete information about caller location, calling identity, devices etc. and can be used to block identity spoofers.